Estimating Likelihood of Directional Drilling Success Through a Practical Application of Projection Uncertainty Models

A key component of managing the risk of directional drilling operations is estimating how closely a well plan can be drilled. Previous work (Willerth & McGregor, 2024) provides a theoretical framework for assessing potential deviation from plan for collision avoidance, however, collision avoidance is not the only time distance from plan is of concern. This work takes a practical look at two common directional drilling scenarios: landing a curve in a target formation and steering a horizontal section within a window. A method of using the projection uncertainty framework to assess confidence in a steerable motor assembly is presented. An assumed motor yield is used to estimate a slide ratio meeting directional objectives on a given well plan. Deviation from plan is estimated using a distribution of motor yields at that slide ratio, assuming recovery using up to full slides at the observed yield. A distribution of rotary tendencies is considered to assess impact on determining true yield, and survey placement relative to slide/rotate intervals is used to place recovery actions. The combined impact on remaining within a drilling window is assessed. Actual assemblies are analyzed for capability and the associated confidence in successfully drilling a plan. The effect of actions such as changing survey course length on potential deviation are considered. Beneficial changes to the BHA and well plan that enable iterative improvement are discussed. The BHAs analyzed were often over-specified in terms of directional capability for the plan being drilled. This is likely from drilling engineers being risk-averse to falling behind plan and factors other than directional accuracy being incentivized (such as minimizing slide-drilling). Excessive dogleg capacity leads to greater uncertainty in the true motor yield being produced as there is increased sensitivity to true-slide-length and BHA rotary tendency. This uncertainty extends into estimates of available yield for recovery steering, reducing confidence in drilling a plan. Similar or better confidence in directional success is possible using less aggressive assemblies if appropriate procedures are used. For example, intelligent use of checkshots to verify directional performance, and incorporating strategic tangents to compensate for geologic and directional uncertainty increases confidence even with less aggressive recovery contingencies. As more data is collected, directional risk models can be further updated to enable intelligent experimentation to identify ideal well plans and BHAs for additional drilling objectives such as wellbore quality or drilling the curve and lateral with single assemblies. Previous work provides a theoretical framework for estimating drilling projection uncertainty for collision avoidance risk assessments. This study extends these efforts to cover how this framework can be defined and updated in a practical way to analyze additional common directional drilling concerns.